Skip to content

AI-native ERP

Projects and guides

AI-native ERP

Start here

Overview First result Persistent backend

Setup

Own provider keys

Product

Product workflows Architecture Design decisions Design and specifications

Reference

Engineering Functions and modules Kernel API Release status Third-party material

Farm-to-restaurant route planner

Start here

Guides

Example data Recorded Barcelona example Routing architecture Routing services Native preview services

Page context for coding agents

Start here

Guides

Capture architecture Provider configuration Extraction decisions Export scope Default provider models Dis Dat — Capture Architecture & Honest Fidelity The deep-grab engine Dis Dat — THE Capability Set (locked)DisDat capture → delivery v2 — executable spec Dis Dat — Design Tokens (the Pro product) · spec v0 Full-site capture playbook Dis Dat — Full-Fidelity Ground-Truth Capture & Handoff (build plan)Grabbing a live WebGL effect's *actual compiled code* and self-hosting it Dis Dat — WebGL reproducibility: the honest ceiling + capture spec

Agent workspace

Start here

Architecture and reference

Source and dependencies Architecture Spending and accounting Engineering workflow Historical records OpenAI agent setup Decision 0013: Managed agent core Agent workspace framework Harness mechanics Add a vertical

Shop records from WhatsApp

Start here

Guides

Product versions and sources THE QR SPEC — the mark on the packing paper, and reading it back Android setup Architecture 0001 · The repository, GitHub flow, releases from tags 0002 · The language model is always Anthropic's 0003 · Paper is read by Anthropic vision; Parsewise is retired from the product 0004 · Artwork is painted by Ideogram, in the library's own style 0005 · One codebase, two builds; the kit runtime stays theatre 0006 · The kit runtime runs in production, behind kit.soyrocoto.com/api 0007 · `main` is protected, and the release tag is the approval 0008 · Three environments: demo and staging deploy from main, production from a tag 0009 · The bridge repository split 0010 · Three repositories: the product, the company canon, the marketing site 0011 · Establish intent and implementation separately Engineering decisions UX decisions First result Source and data provenance Own-account setup Specifications and reference Capability and proof status Workflow and approvals

Renovation marketplace

Start here

Overview First result

Setup

Setup and credentials Database and access

Product

Product workflows UX decisions Architecture

Reference

Functions and routes Release status Third-party material

backend/brain/init.py

All modules · Source

Order-brain: grounded WhatsApp/voice order intake.

Source document ↗

made by atelier neyra

Search documentation

Search words

Type to find a guide or section.

↑ ↓ to choose · Enter to open · Esc to close